Does Microsoft know what IT shops really need?

David Berlind's Feature heap won't undo LAMP's toll on Microsoft brings to mind the importance of understanding the client's needs.

The Microsoft juggernaut came about, in large part, because Bill Gates envisioned a one-stop-shopping model by which the consumer could go to one vendor and buy one product (a PC configured with Windows and Office) and meet 95% of the their needs. No muss, no fuss! You literally need to know nothing about using a computer to go out and select a decent machine which can do everything the typical end user could ever want to do.

The "feature list" is the only meaningful measure that the typical consumer has to go by to determine the "best bang for the buck" and, in the end, the consumer often buys a great deal more features than they end up needing.

The enterprise doesn't care about features. To the enterprise, "best bang for the buck" is measured in productivity as compared to TCO.

As David points out, the Unix/Linux model is built upon modularity -- its greatest strength is its ability permit the IT professional to use every last cycle for a very specific task. Thus, every Unix/Linux implementation has essentially one feature -- the one task it was configured to do.

Until Microsoft recognizes that its enterprise customers are the IT guys working in a corporate "cost center" putting together high-performance servers on a deadline and not the VP down the hall, they are going to face strong competition from Unix/Linux vendors.

I have long maintained that the upfront cost-savings of Linux (over Windows) are largely offset by added personnel and training costs; but if one requires more hardware to do the same task because the OS is too feature-rich, the flexibility afforded the IT professional to strip out those unneeded features in a Unix/Linux implementation becomes somewhat more compelling.

Microsoft's greatest strength in the eyes of the consumer may very well be its greatest weakness in the eyes of the CIO.